This project was developed in the context of the “Artificial Zurichness” research. Deep Learning and Generative Adversarial Networks (GAN) were deployed as a project tool to capture and maintain the essence (genius loci) of a specific site – Zürich – while modifying the type of architecture through additions or variants. From here, a GAN image was selected, and after the classification and analysis process, several concepts that characterize Zurich were extracted: depth, material, minimalism, silhouette and skyline to which the fold was added.
 |
 |
 |
The analysis focused on roofs, openings and texture which led to Zurichness. For the 3D modeling, existing elements were taken and applied with different rules such as light roof openings and changes on volumes. This was combined with minimalism, depth and skyline as Zurich conceps. The fold was used to enhance depth, thereby creating a minimalist effect, and changes of volumes gave to building two potential sections and skylines.
 |
 |
 |
 |
 |
 |
Different levels of privacy were introduced to create an open space that can modulate its privacy with fabrics for bedroom use, and closed spaces for special programs, such as playing or studying. The ground floor becomes a flexible space for possible exhibitions, markets or meetings. The top floor of the tower opens to the public.
Responding to the problem of light, windows cover 5m of illuminated space. As the building has a width of 11m, floor-to-ceiling skylights are used, in addition to taking advantage of the skylight of the neighboring building to open more light points.
The walls of the building are structural, so a unidirectional system will be used as a construction method. The detail represents a section per structural wall, one per non-structural wall and detail with the ground.
Keywords: GAN, Artificial Design, Zurich, Housing
Master project student: Alba Mulero Manzanero
Studio tutors: Christina Doumpioti, Christoph Holz, Mikhael Johanes, Frederick Kim, Gianna Ledermann
Professor: Jeffrey Huang
Image credits: Alba Mulero Manzanero, MxD